Voice source cepstrum coefficients for speaker identification

We propose a novel feature set for speaker recognition that is based on the voice source signal. The feature extraction process uses closed-phase LPC analysis to estimate the vocal tract transfer func-tion. The LPC spectrum envelope is converted to cepstrum coef-ficients which are used to derive the voice source features. Un-like approaches based on inverse-filtering, our procedure is robust to LPC analysis errors and low-frequency phase distortion. We have performed text-independent closed-set speaker identification experiments on the TIMIT and the YOHO databases using a stan-dard Gaussian mixture model technique. Compared to using mel-frequency cepstrum coefficients, the misclassification rate for the TIMIT database reduced from 1.51 % to 0.16%when combined with the proposed voice source features. For the YOHO database the mis-classification rate decreased from 13.79 % to 10.07%. The new fea-ture vector also compares favourably to other proposed voice source feature sets